Calibrating
optical stress signals
for characterizing
10 Gb/s optical
transceivers
Application Note
What is optical stress? (Example 10 GbE)
Figure 1. The N4917A with the option E01, the Receiver Stress Conditioning Unit 10 GBASE-LR/-ER, 10 G Fibre channel and
automation software (CD is not shown)
The N4917A Optical Receiver Stress Test solution with the
conditioning unit N4917A-E01 as shown in figure 1 targets
the standards for 10 GbE [1] and 10 GFC [2].
Jitter histogram
(at waveform
average, may
not be at waist)
Vertical
eye-closure
histograms
(at time-center of eye)
Figure 2 illustrates an optical stressed signal which has
to be applied to an optical receiver. While such a signal
is applied to the optical input, the bit error ratio at the
receiver’s output has to be below a certain level (typically
1e-12) to be compliant.
Figure 3 shows the requirements of such stress signal
generation from the IEEE 802.3ae standard. The fourth-order
Bessel-Thomson filter is used to create ISI-induced vertical
eye closure penalty (VECP). The sinusoidal amplitude
causes additional vertical eye closure, the sinusoidal phase
modulation represents jitter for the jitter tolerance test
according the jitter mask.
P1
AO
Approximate OMA
(difference of means
of histograms)
OMA
P0
Jitter
VECP = 10x log (OMA/AO)
Figure 2. Definition of the optical parameters
Figure 4 illustrates the whole setup of the Optical Receiver
Stress Test Solution including electrical and optical cabling
and including the 86100C DCA-J as calibration tool.
2
OMA:
Optical Modulation Amplitude, measured in
[μW] (“average signal amplitude“)
ER:
Extinction Ratio, high-level to low-level,
measured in [dB] or [%]
UI:
Unit Interval (one bit period)
LR, SR, ER: Flavors of 10 Gb Ethernet standard for Long
Reach (10 km), Short Reach (300 m), Extended
Reach (40 km)
A0:
Vertical eye opening (“innermost eye opening
at center of eye“) [dBm or μW]
VECP:
Vertical Eye Closure Penalty
Calibrating optical stress signals for characterizing 10 Gb/s optical transceivers
)UHTXHQF\V\QWKHVL]HU
)0LQSXW
&ORFNVRXUFH
‡7KHIRXUWKRUGHU%HVVHO7KRPVRQ
ÀOWHULVXVHGWRFUHDWH,6,LQGXFHG
YHUWLFDOH\HFORVXUHSHQDOW\9(&3
‡7KHVLQXVRLGDODPSOLWXGH
interfererFDXVHVDGGLWLRQDO
H\HFORVXUH
3&65[
7HVWSDWWHUQ
2SWLFDO
attenuator
7HVWSDWWHUQJHQHUDWRU
tr
30$5[
73
30'5[
6WUHVVFRQGLWLRQLQJ
6LJQDOFKDUDFWHUL]DWLRQ
PHDVXUHPHQW
(2FRQYHUWHU
6WUHVVFRQGLWLRQLQJ
WKRUGHU
%HVVHO7KRPVRQ
ÀOWHU
‡7KHVLQXVRLGDOSKDVHPRGXODWLRQ
UHSUHVHQWVORZIUHTXHQF\MLWWHU
$SSOLHGVLQXVRLGDOMLWWHU
SHDNWRSHDNDPSOLWXGH
8,ORJVFDOH
6\VWHP
XQGHUWHVW
6LJQDOFKDUDFWHUL]DWLRQ
PHDVXUHPHQW
2VFLOORVFRSH
5 UI
6LQXVRLGDOJHQHUDWRU
&ORFNLQ
0.15 UI
0.05 UI
&58
3% range
Data in
40 kHz
4 MHz
%
(
5
7
UHI5[
10 LB
0DVNRIWKHVLQXVRLGDOFRPSRQHQWRIMLWWHUWROHUDQFHLQIRUPDWLYH
6WUHVVHGUHFHLYHUFRQIRUPDQFHWHVWEORFNGLDJUDP
Figure 3. IEEE 802.3 ae: 52.9.10.1 Stressed Receiver Conformance Test Block Diagram
N4817A
N4915-004
15442A
N4917A-001 FO cable kit SMF
DUT
For testing
connect to DUT
Optical IN
Electrical OUT
For calibration connect to DCA-J
50 Ohm termination
81490A E/O
N4917A - Stress test
conditioning unit
8164A/B LMS
J-Bert N4903A
81576/7A
attenuator
86100C DCA-J
86105B/C
optical mocule
Figure 4. Test Setup Optical Receiver Stress with the N4917A
Calibrating optical stress signals for characterizing 10 Gb/s optical transceivers
3
How to measure the optical parameters with the 86100C DCA-J
The main optical parameters are given in figure 2.
However the following are also important to note:
► the crossing point must be kept at the 50 % of
the levels
► the total jitter must include:
• Random Jitter (RJ)
• Duty Cycle Distortion (DCD)
• Inter-Symbol Interference (ISI)
Examples for the measurement of all the parameters are
shown in figures 5 to 8. All measurements use the PRBS
pattern 2^11-1 unless otherwise specified.
Extinction Ratio (ER)
The measurement of ER is shown in figure 5. There is a
predefined routine available in the eye diagram mode. There
is an Extinction Ratio Correction Factor (ERCF) that must be
applied for measuring large ER as discussed in [3].
Figure 5. ER & crossing point by Eye Diagram Mode
(PRBS pattern, ERCF enabled)
Crossing Point
The eye diagram mode also provides a routine for the
measurement of the crossing point.
Optical Modulation Amplitude (OMA)
The OMA is measured as shown in figure 7. For this measurement a ..0011.. pattern is used. The OMA measurement
is equivalent to the Eye Amplitude routine available in the eye
diagram mode. By using this pattern and the Eye Amplitude
routine this is equivalent to the definition as shown in
figure 2.
Figure 6. TJ (BER 1E-3), RJ, DCD & ISI by Jitter Mode in time
4
Calibrating optical stress signals for characterizing 10 Gb/s optical transceivers
Jitter, A0 and VECP
Jitter and A0 are measured with the Jitter Mode. TJ is
available from the Time Menu as shown in figure 6. A0
is available as Eye Opening from the Amplitude Menu as
shown in figure 8. VECP can not measured directly, it is
calculated by the formula:
VECP = 10* log(OMA/A0).
Requirements
Jitter Measurements in time and amplitude require specific
options (200, 300) of the 86100C DCA-J.
The measurements shown here are taken with help of a
Precision time base module (86107A-010). This is recommended to verify the sub-ps RJ performance. However it
is not a requirement for the calibration of the setup.
Figure 7. OMA by Eye Diagram Mode (..1100.. pattern, Routine: Eye Amplitude)
Sufficient optical power must be provided to the optical
inputs of the DCA-J optical modules (1 mW for 86105B,
0.2 mW for 86105C). This is important for the verification
of RJ jitter.
Figure 8. A0 (BER 1E-3) by Jitter Mode in Amplitude, Parameter: Eye Opening
Calibrating optical stress signals for characterizing 10 Gb/s optical transceivers
5
Configuring the Measurements of the 86100C DCA-J
For the Jitter and A0 measurement there is an important
statement in the IEEE 802.3ae standard saying:
The test signal includes vertical eye closure and
high-probability jitter components. For this test, these
two components are defined by peak values that
include all but 0.1% for VECP and all but 1% for jitter of
their histograms. Histograms should include at least
10,000 hits, and should be about 1%-width in the
direction not being measured. Residual low-probability
noise and jitter should be minimized—that is, the outer
slopes of the final histograms should be as steep as
possible down to very low probabilities. (from: 52.9.9.2
Stressed receiver conformance test signal characteristics
and calibration)
The 86100C DCA-J does not allow to set BER to 1e-2
directly, the lowest BER can be set for 1e-3. So the TJ for
BER 1e-2 is calculated by:
TJ (BER 1e-2) = TJ (BER 1e-3) – 2* RJ
This is derived from the nature of random jitter being
Gaussian distributed. Figure 9 shows the properties of the
Gaussian distribution. Figure 10 shows the histogram of a
jitter measurement excluding x % of the events.
The Gaussian distribution helps to describe something
which is basically infinite like random jitter.
Basically this means that random jitter during the calibration
of the stressed eye should be eliminated. The definition of
the peak values can be mapped to the capabilities of the
86100C DCA-J by defining the TJ and Eye Opening value as:
TJ based on (BER 1e-2),
A0 (and VECP) based on (BER 1e-3)
(OMA is not dependent on BER)
mean value
Normalized
events
sigma
sigma
sigma
sigma
time
Figure 9. Gaussian distrubution and the relation of BER = 1 – # events (%)/100
6
Calibrating optical stress signals for characterizing 10 Gb/s optical transceivers
Figure 11. ISO BER measurement from a Bit Error Ratio Tester (BERT)
x% of events
x% of events
(100 – 2x)%
of events
Figure 10. Histogram measurement with excluding x % of the
measured events
We‘d like to give a more intuitive explanation here, see [4]
for a more scientific explanation:
The distribution is described by the mean value and
the standard deviation sigma. As more of the events are
included the more times sigma has to be included:
# events = n * sigma
The table in figure 9 lists the relation with n.
On the other side this can be expressed as a BER threshold.
The relation of BER can be expressed as:
Figure 12. 86100C DCA-J Jitter measurement configured for BER 1e-3
BER = 1 - # events (%) / 100
From the table in figure 9 it gets obvious that the difference
between TJ (BER 1E-2) and TJ (BER 1E -3) is 2 x sigma.
In this case sigma is the reading for RJ.
Figure 11 shows the BER contour measurement. This is a
measurement available on a Bit Error Ratio Test System
(BERT). It shows the eye contour as a function of BER. The
eye width/amplitude can be directly read according to the
desired BER threshold. In this example the color coding on
the right side is the key for the BER threshold in the picture.
The eye width or the corresponding jitter and the eye
amplitude as a function of BER can also be measured with
the 86100C DCA-J with the jitter measurements. Figure 12
shows the jitter measurement in time while the DCA-J is
configured for BER 1E-3.
Calibrating optical stress signals for characterizing 10 Gb/s optical transceivers
7
Dependencies of Parameters
Pattern Dependency
The IEEE802.3ae defines a bandwidth limitation within the
stress conditioning. The bandwidth is set to 0.75 of the data
rate. This ensures the amplitude of a single bits does not
achieve the 100% amplitude level. This introduces ISI type
of jitter.
The standard requires to use a Bessel Thompson Filter
which is also called a Gaussian filter. Compared to filter
types like Butterworth or Chebyshev, a Gaussian filter
provides a clean step response, see Figure 13.
Gaussian
Jitter and Pattern Dependency
Butterworth
Chebyshev
The individual jitter parameters, as discussed in the following
paragraphs, are measured with the DCA-J as shown in
Figure 6.
The Figures 14 & 15 show the influence of the pattern on the
total jitter and its subcomponents. The patterns used are:
PRBS 2^7-1
PRBS 2^11-1
PRBS 2^15-1
clock pattern ( .. 0011 ..)
K28.5
CJTPAT
The measurements were done at a data rate of 10.3125 Gb/s.
The plots in both figures compare:
The Total Jitter (TJ)
Inter-symbol Interference (ISI)
Random Jitter (RJ)
Periodic Jitter (PJ)
Duty Cycle Distortion (DCD).
The curves compare a pure electrical filter of 7.46 GHz
bandwidth with Bessel-Thompson characteristic (called
electrical filter, see Figure 13) and the N4917A setup once
without any jitter injection (called optical) and once with
jitter injection of sinusoidal interference = 100 mV @ 1GHz
and PJ = 100 mUI @ 4 MHz (called optical w jitter).
8
Figure 13. Example of a Bessel-Thompson Filter (Gaussian)
with cut-off at 7.46 GHz and the step response compared to
Butterworth/Chebyshev filter type
Results:
Figure 14:
• the ISI is pattern dependent, N4917A is very similar
to the pure filter,
• TJ follows ISI behavior while the jitter modulation
affects the total jitter as expected
Figure 15:
• RJ & DCD are not pattern dependent, N4917A is very
similar to pure filter.
• PJ is not pattern dependent, N4917A w/o jitter
modulation is marginally higher than the pure filter,
with jitter modulation PJ behaves as expected.
Conclusion:
• The specifications of filter and pattern create a
specific ISI content in the total jitter budget
• The use of alternate patterns for compliance tests
must be avoided
• The other jitter components beside ISI can be
added flexibly and calibrated at any pattern
Calibrating optical stress signals for characterizing 10 Gb/s optical transceivers
Total jitter (TJ) & ISI jitter
45
40
35
TJ electrical filter
ps
30
TJ optical w/o jitter
25
TJ optical w jitter
20
ISI electrical filter
15
ISI optical
ISI optical w jitter
10
5
CJ
TP
AT
.5
28
K
clk
00
10
2^ 1
-51
2^ 1
-11
PR
BS
2^ 7
-1
0
pattern
Figure 14. Total Jitter (TJ) and Inter-Symbol Interference (ISI)
RJrms, PJ & DCD jitter
12
RJ electrical filter
10
RJ optical w/o jitter
RJ optical w jitter
8
ps
PJ electrical filter
6
PJ optical
PJ optical w jitter
4
DCD electrical filter
DCD optical
2
DCD optical w jitter
TP
AT
CJ
.5
28
K
clk
00
10
2^ 1
-51
2^ 1
-11
PR
BS
2^ 7
-1
0
pattern
Figure 15. Random Jitter (RJ), Periodic Jitter (PJ) and Duty Cycle Distortion (DCD)
Calibrating optical stress signals for characterizing 10 Gb/s optical transceivers
9
Validation of the methodology used
The following is a key recommendation in the
IEEE 802.3ae standard:
A short pattern may be used for calibration if the
conditions described in 52.9.9.1 are met, but increases
the risk that the longer test pattern used during testing
will overstress the device under test. In any case, a
pattern shorter than PRBS10 is not recommended.
(from: 52.9.9.2 Stressed receiver conformance test
signal characteristics and calibration)
The impact of using longer patterns
The question remains:
how much jitter is added when switching from
PRBS 2^11-1 to 2^31-1?
Figure 16. Histogram Measurement of the optical signal with a PRBS 2^11-1
This can be measured with the 86100C in Eye/Mask mode
with help of the histogram measurement. The results are
shown in figure 16 & 17 as p-p values, see the red circles.
The difference is measured as 1.33 ps for the N4917A
optical signal (Figure16 PRBS 2^11-1 p-p = 23.67 ps,
Figure 17 PRBS 2^31-1 p-p = 25 ps, difference = 1.33 ps).
Conclusions:
Based on the standard’s recommendation and the measurement capabilities of the 86100C DCA-J the PRBS 2^11-1 is
used for the calibration of the N4917A.
The use of shorter PRBS pattern for calibration creates a
marginal impact.
Figure 17. Histogram Measurement of the optical signal with a PRBS 2^31-1
10
Calibrating optical stress signals for characterizing 10 Gb/s optical transceivers
Data Rate Dependency
Figure 18 & 19 show the linearity of ER and OMA over the
data rate.
Conclusions:
A calibration has to be performed at the actual data rate
A change of the data rate after performing a
calibration must be avoided
data rate/Gb/s
Figure 18: ER vs. data rate, linearity
.6
11
.3
11
11
.7
.4
10
.1
10
9.8
86105C
Delta OMA/%
9.5
.6
11
.3
Linearity OMA
11
11
.1
10
.4
10
.7
10
9.8
Delta ER/%
9.5
Linearity ER
10
•
•
86105C
data rate/Gb/s
Figure 19: OMA vs. data rate, linearity
Calibrating optical stress signals for characterizing 10 Gb/s optical transceivers
11
Dependencies of Optical Avg. Power,
ER, OMA & A0
Figure 21 shows the relationship of ER, OMA and optical
power to the attenuator setting: ER is not affected by
the attenuator setting, while OMA and optical power is
a linear function of the attenuation.
Figure 20 shows the dependency of OMA, A0 and optical
average power as function of ER. As ER depends on
the electrical amplitude of the data signal, all the
optical parameters depend directly on the amplitude of
the electrical data signal. Figure 20 includes a 3-point
extrapolation for OMA and A0. 3 points from the
measured curve are taken (referenced as Calculate
A0/OMA) and used to extrapolate the curve
(referenced as 3 point cal A0/OMA).
Conclusions:
The deviation of the approximation of ER, OMA and A0 is
marginal, so this is proven to be a valid algorithm for the
calibration.
The attenuator does not need a calibration. A single
power measurement as a reference is sufficient
OMA & AO @ at fixed attenuator settings of 2 dB
3
2.5
OMA
AO
Pavg
3 point cal OMA
3 point cal AO
calculate AO
Calculate OMA
mW
2
1.5
1
0.5
0
0
2
4
6
8
10
12
ER/dB
Figure 20. Optical avg. Power, MA and AO as a function of ER at fixed attenuator
avg power, ER & OMA @ amp = 250 mV vs. Attenuator setting
5
0
–5
result
power/dBM
ER/dB
OMA/dBM
–10
–15
–20
0
2
4
6
8
10
12
14
attenuator setting
Figure 21. Optical avg. Power, OMA & ‚ER as a function of attenuator setting
12
Calibrating optical stress signals for characterizing 10 Gb/s optical transceivers
Dependencies of Periodic Jitter (PJ)
and Sinusoidal Interference (SI)
Figure 22 to 24 are examples of histogram measurements
of the optical signal with various SI & PJ modulation.
Figure 22:
PJ = 0 ps
SI = 0 mV
TJ = 19.8 ps
A0 = 372 uW
VECP = 2.66
Figure 23:
PJ = 20 ps
SI = 0 mV
TJ = 37.6 ps
A0 = 354 uW
VECP = 2.87
Figure 24:
PJ = 0 ps
SI = 100 mV
TJ = 37.8 ps
A0 = 220 mW
VECP = 4.94
Calibrating optical stress signals for characterizing 10 Gb/s optical transceivers
13
How to get numbers for TJ, A0 & VECP
Conclusions:
It should be noted that TJ, A0 and VECP are dependent on
both SI and PJ. So it is impossible to provide numeric entries for TJ, A0 and VECP and convert this straight forward
into numbers for SI and PJ.
•
As the relationship of A0 and VECP versus SI and
PJ is mostly linear, this can be modeled with a
2-point approximation
•
This is similar for the total jitter, this depends also
on SI and PJ programming
However it is possible to calculate TJ, A0 and VECP from
numeric entries for PJ and SI values.
Figure 25 and 26 show the relation of A0 and VECP as a
function of Sinusoidal Interference (SI) programming and
periodic jitter (PJ) programming.
VECP vs. SI & PJ
A0 vs. SI & PJ
400
7.00
350
6.00
300
5.00
var SI
var PJ
200
VECP
AO
250
4.00
var SI
var PJ
3.00
150
2.00
100
1.00
50
0.00
0
0
10
20
30
Jitter/ps
Figure 25. A0 as a function of SI and PJ
14
40
50
60
0
10
20
30
Jitter/ps
40
Figure 26. VECP as a function of SI and PJ
Calibrating optical stress signals for characterizing 10 Gb/s optical transceivers
50
60
Summary on Calibration Conclusions
• Pattern affects TJ/ISI. The IEE 802.3 ae standard
recommends a calibration using a PRBS > 2^10-1. It
could be shown that the use of 2^11-1 is acceptable and
can be effectively used with the automated routines of
the 86100C DCA-J.
• The data rate affects all parameters significantly.
The calibration must be performed at a given rate.
• ER, OMA, A0/VECP and total jitter depend on electrical
amplitude, optical attenuation and the programming of
SI and PJ. The calibration can be performed with help
on look-up tables based on linear approximation.
The N4917A Calibration Software
The Agilent N4917A optical receiver stress test calibration
and automation software takes these findings into account.
It automatically calibrates the optical stress signal with
the DCA-J according to a list of selectable standards and
with user adjustable parameters. Additionally it allows
automated receiver sensitivity measurements, such as
BER versus OMA.
Calibration Modes
1. At fixed data rate and compliance point of standard
specific Compliance Mode: ER, OMA, fixed jitter
for meeting compliance window, except variable
PJ (50 .. 150 mUI requested), fixed pattern
(PRBS 2^11-1 for cal, 2^31-1 for measurement)
2. At fixed data rate of standard with variable ER, OMA,
SI & PJ jitter standard specific Manual Mode: with
reading for TJ, A0 and VECP depending on ER, OMA,
SI and PJ parameter setting, fixed pattern
(PRBS 2^11-1 for cal, 2^31-1 for measurement)
3. At non-standard data rate with variable ER and OMA,
SI & PJ Custom Standard: with variable data rate &
wavelength, ER, OMA, SI and PJ parameter setting,
fixed pattern (PRBS 2^11-1 for cal, 2^31-1 pattern for
measuremen), no reading for TJ, A0 and VECP
Calibration steps performed by the N4917A
Step 1: instrument check
• check for required Models & Options
Step 2: cabling check
• check for correct electrical and optical cabling
Step 3: instrument preparation
Scope User Cal, E/O optimization for DCD
• cabling optimization
• optimization of intrinsic ISI due to cable
tolerances
Step 4: parameter calibration, reference point
measurements for the look-up table for:
• 3-point extrapolation for ER, crossing point,
OMA, A0, TJ
• 2-point extrapolation for SI and PJ on A0 and TJ
Calibrating optical stress signals for characterizing 10 Gb/s optical transceivers
15
Implementation of the signal conditioning
The channel impairment
Figure 28 to 31 explain why it is hardly possible to implement
the signal conditioning as simply as the standard states. It
is shown that a signal compensation is needed to create
the stress signal as requested by the standard. Figure 28
shows a typical S21 curve of an optical modulator. The
relationship between attenuation and frequency is not
linear, the higher the frequency the higher the attenuation
(the losses). This looks pretty similar to the characteristics
of a PC Board used at gigabit speed. This behavior creates
especially ISI type jitter as can be seen in the screen shot
of figure 30.
The standard requests signal conditioning using a BesselThompson filter and an adder for sinusoidal interference
(as shown in figure 29). When combining such a modulator
with the simple conditioning circuit the resulting figures for
intrinsic jitter and VECP are shown in figure 31. The result
is a TJ which is already exceeding the compliance limit and
a VECP which comes close to the limit without required
adding of PJ and SI type of jitter for the compliance point.
a1
b1
b1
b2
S11 S12
= S S
21
22
–10
–12
–14
–16
dB
–18
–20
S21 [dB] @ –13 dBm
–24
S21 [dB] @ –7 dBm
–26
S21 [dB] @ –5 dBm
–28
S21 [dB] @ –3 dBm
–30
0.0E + 00
1.0E + 10
2.0E + 10
3.0E + 10
4.0E + 10
5.0E + 10
Frequency/Hz
Figure 28. The S21 Parameter of an optical modulator (attenuation vs. frequency)
16
a1
a2
Figure 27. S-(or scattering) parameters describe the input (S11; S22)
and the transmission (S21, S12) behavior of a two or multi-port device.
They are complex figures and depend on frequency. The graph in Figure 28
shows just the magnitude. For more information on S-parameters see [5].
S21 of Modulator
–22
b2
a2
2-port
network
Calibrating optical stress signals for characterizing 10 Gb/s optical transceivers
SI from J-BERT N4903A
Filter
to E/O
J-BERT N4903A
data_out
Figure 29. A simple conditioning circuitry with filtering and SI modulation
Figure 30. A measurement of ISI/TJ using the conditioning circuitry from
Figure 29 together with an optical modulator with the typical S21 parameter
from Figure 28
Calibrating optical stress signals for characterizing 10 Gb/s optical transceivers
17
TJ @ 1310 nm & 10.51875 Gb/s
UI
.2 UI
0.03
0.25
0.20
0.15
0.10
0.05
0.00
TJ @ 1310 nm &
10.51875 Gb/s
0
5
10
15
ER/dB
VECP @ 1310 nm & 10.51875 Gb/s
2.5
.2.2 dB
dB
2.0
VECP @ 1310 nm
& 10.51875 Gb/s
1.5
1.0
0.5
0.0
0
5
10
ER/dB
15
ER = 3.5 dB
Figure 31. The TJ and VECP as function of ER of the conditioning circuitry from Figure 29 together with an
optical modulator with the typical S21 parameter from Figure 28
18
Calibrating optical stress signals for characterizing 10 Gb/s optical transceivers
Signal Conditioning including
compensation of ISI
The solution to this behavior is adding a compensation into
the stress conditioning circuitry which compensates for the
losses of the cabling and the S21 of optical modulator while
keeping the desired stress of the bandwidth limitation by
the Bessel-Thomson filter. The resulting waveforms with
the built-in compensation are given in Figure 32 & 33.
Figure 32 is the signal without any added jitter, Figure 33
adds PJ and SI jitter to compliance point requirements.
The built-in compensation is designed in such a way that
the resulting ISI of the optical signal resulting from the
whole electro-optical setup equals mostly the ISI obtained
from a Bessel-Thompson filter with the electrical bandwidth
required by the standard. The ISI values in Figure 14
(ISI = 10 ps) and Figure 32 (ISI = 10.4 ps) shall be compared.
Figure 32. The jitter measurement on the optical signal of the N4917A setup
with no jitter injection
Final Conclusions:
The 86100C DCA-J is an effective tool for calibrating the
optical parameters, it offers all necessary automated
routines . The configuration of the automated routines
is easy to make.
The dependency of parameters is a little more complex.
There is some data pattern dependency, but it could be
shown that using the PRBS 2^11-1 for calibration has a
marginal compared to effect for measuring with a PRBS
2^31-1 which the standard requires. The data rate affects
the parameters significantly, so this needs a calibration at
the actual rate.
The implementation of the stress conditioning N4917A-E01
is not as easy as one might believe from the standard‘s
definition. It was shown that a compensation is needed
for maintaining proper stress condition according to the
standard‘s requirements of adding stress. It could be shown
that this is possible to achieve it to exactly the stress as
required by the standard.
The parameters affecting Jitter and VECP are calibrated
with help of 2- and 3-point extrapolation. The accuracy
achieved is better than 10% at the compliance point
This was verified using a set of two different BERTs, three
different stress conditioning units, three different E/O
converters and three different optical scope modules in
a random combination.
Figure 33. The jitter measurement on the optical signal of the N4917A setup
with SI and PJ jitter injection according the compliance requirements
N4917A Optical Receiver Stress Test Set
Photonics Tech Briefs
Product of the Year 2007.
Selected by readers of Photonics Tech Briefs as readers
choice product of the year.
Calibrating optical stress signals for characterizing 10 Gb/s optical transceivers
19
References
Related Literature
[1] IEEE Std 802.3ae™-2002
• Agilent N4917A Optical Receiver Stress Test Solution,
Data Sheet, Literature Number 5989-6315EN
[2] Fibre Channel — 10 Gigabit (10GFC), American National
Standard for Information Technology
[3] Improving the Accuracy of Optical Transceiver
Extinction Ratio Measurements, Agilent Technologies
Application Note 1550-9, 2005, 5989-2602EN
[4] Digital Communications Test and Measurement,
Dennis Derickson & Marcus Müller,
Prentice Hall 2007, ISBN13: 978-0-13220910-6
[5] Agilent Technologies, S-Parameter Design,
Application Note AN 154, 2000/2006, 5952-1087
• Agilent J-BERT N4903A High-Performance Serial BERT
with Complete Jitter Tolerance Testing, Data Sheet,
Literature Number 5989-2899EN
• Agilent 86100C Wide-Bandwidth Oscilloscope Mainframe
and Modules Technical Specifications,
Literature Number 5989-0278EN
• Agilent Technologies 81490A Reference Transmitter
Technical Specifications, Literature Number 5989-7326EN
• Agilent 81495 Reference Receiver Technical Specifications,
Literature Number 5989-7526EN
Web Link
www.agilent.com/find/n4917
20
Calibrating optical stress signals for characterizing 10 Gb/s optical transceivers
www.agilent.com
Remove all doubt
Our repair and calibration services will get your equipment
back to you, performing like new, when promised. You will
get full value out of your Agilent equipment throughout
its lifetime. Your equipment will be serviced by Agilenttrained technicians using the latest factory calibration
procedures, automated repair diagnostics and genuine
parts. You will always have the utmost confidence in
your measurements.
Agilent offers a wide range of additional expert test
and measurement services for your equipment, including
initial start-up assistance onsite education and
training, as well as design, system integration, and
project management.
For more information on repair and calibration
services, go to
www.agilent.com/find/removealldoubt
www.agilent.com/find/emailupdates
Get the latest information on the products and
applications you select.
www.agilent.com/find/agilentdirect
Quickly choose and use your test equipment solutions
with confidence.
www.agilent.com/find/open
Agilent Open simplifies the process of connecting and
programming test systems to help engineers design,
validate and manufacture electronic products. Agilent
offers open connectivity for a broad range of system-ready
instruments, open industry software, PC-standard I/O and
global support, which are combined to more easily integrate
test system development.
For more information on Agilent Technologies’ products, applications
or services, please contact your local Agilent office. The
complete list is available at:
www.agilent.com/find/contactus
Americas
Canada
Latin America
United States
(877) 894-4414
305 269 7500
(800) 829-4444
Asia Pacific
Australia
China
Hong Kong
India
Japan
Korea
Malaysia
Singapore
Taiwan
Thailand
1 800 629 485
800 810 0189
800 938 693
1 800 112 929
0120 (421) 345
080 769 0800
1 800 888 848
1 800 375 8100
0800 047 866
1 800 226 008
Europe & Middle East
Austria
Belgium
Denmark
Finland
France
Germany
0820 87 44 11
32 (0) 2 404 93 40
45 70 13 15 15
358 (0) 10 855 2100
0825 010 700*
*0.125 €/minute
01805 24 6333**
**0.14€/minute
1890 924 204
972-3-9288-504/544
39 02 92 60 8484
31 (0) 20 547 2111
34 (91) 631 3300
0200-88 22 55
0800 80 53 53
44 (0) 118 9276201
Ireland
Israel
Italy
Netherlands
Spain
Sweden
Switzerland
United Kingdom
Other European Countries:
www.agilent.com/find/contactus
Revised: March 27, 2008
Product specifications and descriptions in this document
subject to change without notice.
www.lxistandard.org
LXI is the LAN-based successor to GPIB, providing
faster, more efficient connectivity. Agilent is a founding
member of the LXI consortium.
© Agilent Technologies, Inc. 2008
Printed in USA, June 10, 2008
5989-8393EN